Matching Heuristics

What you have to do:

Options:

Kind of Algorithm:

BOTTLENECK

This algorithm sorts the weightings and then searches binary for the smallest weighting p so that  M = {p(i,j) | p(i,j) >= p} contains a perfect matching. The according operations are inserted into the schedule and the weightings are deleted. This process will be iterated until all operations have been inserted into the schedule.

WEIGHTED

A maximum weighted matching is calculated. The according operation are inserted into the schedule and the weightings are deleted (eg. set to the lowest possible value). This will be repeated until all operations have been inserted into the schedule.

Kind of matching:

 This option allows you to choose which kind of weightings will be used for the matching algorithms.

MIN

All processingtimes will be subtracted from the longest possible processingtime, the result will be used as weighting. That means the algorithm trys to minimize the average processingtime for the inserted operations.

MAX

The processingtime for each operation will be used as weighting. That means the algorithm trys to maximize the average processingtime for the inserted operations.

HEADS

The first set of operations to insert will be calculated as if the MIN parameter has been set. After inserting a set of operations the heads for all remaining operations will be calculated and added to the processingtime for that operation. The result will be substracted from the longest possible processingtime and be used as weighting. That means the algorithm trys to insert a set of operations that way, that the current total completion time will get minimized.